archiv~1.txt: Re: SETI-L: Everybody may be afraid & a Answer

Re: SETI-L: Everybody may be afraid & a Answer

Alfred A. Aburto Jr. ( (no email) )
Fri, 26 Mar 1999 17:00:23 -0800

With a 20MHz bandwidth and 1Hz bins they get an observation
every second! 3 hours of data is therefore almost 1.7 Terabytes
(assuming data is stored as 4 byte real & imaginary components).

I don't know exactly what they do but the above shows that it is
very easy (in a matter of a few hours) to accumulate over a
terabyte of data!

Al

> mbernstein@c0t.org wrote:
>
> I realize Moore's law, and while beowulf clusters are not optimized for
> SETI, they are optimized for what they do. Even 5 terabytes per star
> seems like an amazing amount of data,.,.,I mean I'm not calling them liars
> by any stretch, it's just seems to be way to much data.,,
>
> michael bernstein
> mbernstein@c0t.org
>
> i could understand a few gig.,.,MAYBE even half a terabyte, but even 5
> terabytes? how long on average does it take to scan 1 star?!
>
> On Thu, 25 Mar 1999 MarcusJohn@aol.com wrote:
>
> > In a message dated 3/25/99 8:04:32 PM Eastern Standard Time,
> > mbernstein@c0t.org writes:
> >
> > > I also would like to raise another point. The Seti Institute said that
> > > the reason they do not post data is because for each star they scan they
> > > accumulate 15 terabytes of data. I don't know how many of you are
> > > computer professionals, or even know that much about computers, but that
> > > is an extremely ludicrous number.
> >
> > You fail to appreciate the power of Moores Law. In other words, we as a
> > species are becoming computationally more powerful all the time, at an
> > exponential rate of increase. Never say any (put your favorite computing value
> > here) is ludicrous. Don't forget, Bill Gates once said that no one would ever
> > need more than 640k of memory.
> >
> > Anyway, they generate 5 teras, not 15. And they don't store it any longer than
> > they need to process it. And processing is done in near real time. Here is the
> > link from the Seti Institute describing their data processing rationale:
> >
> > http://www.seti-inst.edu/phoenix/data_archive.html
> >
> > > They said all processing is done in
> > > realtime. If you can show me a computer that can process 15 terabytes of
> > > information into complex search algorithms in the short time they spend to
> > > scan each star then I don't know what I'd do.
> >
> > Well, your only off by a factor of three. Start looking for something that
> > 'you don't know what to do', hehe.
> >
> > > It's impossible at this
> > > point in time, even with an extremely large cluster to process 15
> > > terabytes of data. Moby Dick is approx 200kilobytes in length last time i
> > > checked. that makes each star's data 75,000,000 times the size of moby
> > > dick, which I realize is not ascii text, but is still a number that is way
> > > way off.
> > >
> > > michael bernstein
> > > mbernstein@c0t.org
> > >
> > >
> > > btw, I've used beowulf clusters on 1terabyte raid systems, and it has
> > > taken WEEKS to do certain things. I'm sure the seti algorithms are
> > > equally complex.
> >
> > Your Beowulf clusters are not optimized for seti work. So the analogy may not
> > be so valid.
> >
> > I know it sounds incredulous, but personally, I take the Seti Institute at
> > their word.
> >
> > John Marcus.
> >